Radar Imaging Using a Wideband Adaptive Array

The focus of this paper is to investigate the application of wideband adaptive array processing techniques to the problem of radar imaging. In particular we are interested in joint range-angle estimation with angular resolution improvement for small, relative to A , antennas. The simulations and experimental results indicate that this approach is viable in a practical sense, and yields significant angular resolution improvement over conventional methods. The approach uses conventional Fourier techniques for downrange information and uses adaptive beamforming for the azimuth dimension. The basic approach is to transmit a wideband set of continuous wave signals, then apply spatial resampling to the received data to correct for the fixed element spacing, then Fourier transform this data to extract range information. We then have a spectral estimation problem at each range cell. By using data associated with each range bin, the angular spectrum is computed using a minimum variance spectral estimate. In general, adaptive array theory is based on the narrow band requirement that the array aperture size is much less than the inverse relative bandwidth. This implies that plane waves are parameterized primarily by their angle of arrival. For this work the narrowband assumption is not valid since we will be using bandwidths of at least 20% of the carrier frequency. Spatial resampling can estimate the array data that would occur if the antenna spacing were varied physically as a function of frequency. See [2] for a discussion of spatial resampling applied to underwater acoustics. Another problem that must be addressed is that parametric spectral estimation methods require a covariance matrix to be estimated from multiple, uncorrelated snapshots of the array output. We introduce a method that amounts to an induced Doppler shift that can be used to generate the required data necessary for angular spectral estimates. The final result is a range-angle plot of backscattered energy.